Inhibition of SRC expression and activity inhibits tumor progression and metastasis of human pancreatic adenocarcinoma cells in an orthotopic nude mouse model

Abstract

The nonreceptor protein tyrosine kinase Src is overexpressed in 70% of pancreatic adenocarcinomas. Here, we describe the effect of molecular and pharmacological down-regulation of Src on incidence, growth, and metastasis of pancreatic tumor cells in an orthotopic model. Src expression in L3.6pl human pancreatic tumor cells was reduced by stable expression of a plasmid encoding small interfering RNA (siRNA) to c-src. In stable siRNA clones, Src expression was reduced >80%, with no change in expression of the related kinases c-Yes and c-Lyn, and proliferation rates were similar in all clones. Phosphorylation of Akt and p44/42 Erk mitogen-activated protein kinase and production of VEGF and IL-8 in culture supernatants were also reduced (P < 0.005). On orthotopic implantation of varying cell numbers into nude mice, tumor incidence was unchanged; however, in the siRNA clones, large tumors failed to develop, and incidence of metastasis was significantly reduced, suggesting that c-Src activity is critical to tumor progression. To examine this possibility further, animals bearing established wild-type tumors were treated with the Src/Abl-selective inhibitor BMS-354825 (dasatinib). Tumor size was decreased, and incidence of metastases was significantly reduced in treated mice compared with controls. These results demonstrate that Src activation contributes to pancreatic tumor progression in this model, offering Src as a candidate for targeted therapy.

Figure 1

c-Src siRNA causes a specific reduction in c-Src expression without affecting in vitro growth rates. A: Stable G418-resistant L3.6pl clones expressing either c-src-targeted siRNA (siSrc C1, C2, and C3) or vector controls were generated from parental L3.6pl cells as described in Materials and Methods. Fifty micrograms of RIPA cell lysates were resolved by 8% SDS-PAGE and Western blot analysis performed using anti-Src monoclonal antibody 327, anti-Yes, or anti-Lyn antibodies. Membranes were stripped and re-probed for β-actin as a loading control. Blots are representative of three experiments. B: siSrc C1, C2, and C3; parental L3.6pl cells; or cells expressing empty vector were plated on day 0 as described in Materials and Methods. 3-(4,5-Dimethyl-2-thiazol-2-yl) 2,5-diphenyltetrazolium bromide assays were performed 1, 24, 48, 72, and 96 hours after plating. Values represent the mean of three experiments.

Figure 2

Reduced c-Src expression results in decreased activation of Akt and Erk 44/42. Parental L3.6pl cells or clonal variants expressing empty vector or c-src-targeted siRNA (siSrc C1 and C2) were plated and maintained as described in Figure 1. Twenty-four hours after plating, cells were serum-starved for 24 hours, and cell lysates were harvested as described in Materials and Methods. Expression of phosphorylated Akt and Erk 44/42 was determined by Western blot analysis. Blots were stripped and re-probed for total Akt and Erk 44/42, respectively. Anti-vinculin Western blot analysis was performed as a loading control. Blots are representative of three independent experiments.

Figure 3

siRNA-mediated c-Src knockdown results in decreased expression of IL-8 and VEGF. siSrc C1, C2, and C3; parental L3.6pl cells; and vector controls were plated and maintained as described in Materials and Methods. Twenty-four hours after plating, the cell culture medium was replaced with serum-free medium and allowed to incubate for an additional 24 hours, at which point cell lysates and culture supernatants were harvested. IL-8 and VEGF levels were measured by ELISA as described in Materials and Methods and expressed relative to total cellular protein (picograms/milligram). Expression levels of both IL-8 (black bars) and VEGF (gray bars) were reduced significantly (P < 0.005, n = 3) relative to controls.

Figure 4

siRNA-mediated reduction in c-Src expression is maintained in tumor cells grown in vivo. siSrc clones, parental L3.6pl cells, and vector controls were injected into the pancreases of nude mice (5 × 10⁵ cells/mouse) on day 0. After 6 weeks, tumors were harvested and frozen in liquid nitrogen. Tumors were homogenized in RIPA, tumor cell lysates were quantitated, and 50 μg of total protein was resolved by 8% SDS-PAGE. Blots were probed for total c-Src and Lyn as described in Materials and Methods. Because of low expression of c-Yes, 500 μg of tumor lysates was immunoprecipitated with the anti-Yes antibody, followed by anti-Yes Western blot analysis. Anti-β-actin Western blot analysis was performed as a total protein loading control. Results are representative of at least three independent Western blots. A specific reduction in total c-Src levels was observed in the siRNA clones after 6 weeks growth in vivo in the absence of G418.

Figure 5

Src expression is decreased in tumors produced from siRNA-expressing clones. L3.6pl parental, vector, and siSrc tumor samples were harvested and prepared for staining as described in Materials and Methods. Presence of tumors was confirmed by hematoxylin and eosin staining (top panels); Src expression was examined by immunohistochemistry (middle panels) and immunofluorescence (bottom panels) as described in Materials and Methods. The results displayed in the figure are representative of sections of all tumors obtained from L3.6pl parental (left panels), vector-transfected (middle panels), and siSrc (right panels; siSrcC1 pictured) tumors, in which Src was reduced by expression of an siRNA as described in Materials and Methods.

Figure 6

Comparison of c-Src expression in primary tumors and liver metastases. L3.6pl parental and siSrc primary tumor and liver samples (with and without metastases) were harvested and prepared for staining or Western blot analysis as described in Materials and Methods. A: Hematoxylin and eosin staining was performed to verify the presence of tumors and metastases, compared with uninvolved tissue. Src expression was examined by immunofluorescence. All staining was performed as described in Materials and Methods. For liver metastases, a comparison of Src staining in metastasis (M) is shown relative to uninvolved liver (N). The images shown are representative of sections obtained from at least five tumor samples. Results are shown for siSrc C1. B: Anti-Src Western blot analysis was performed on lysates of L3.6pl parental primary tumors, liver metastases, and normal liver as described in Materials and Methods.

Figure 7

Phospho-Akt, phospho-Erk 44/42, and CD31⁺ vessel infiltration are reduced in tumors from siRNA-expressing clones. L3.6pl parental and siSrc tumor samples were harvested and prepared for staining as described in Materials and Methods. Tumor samples were stained with phosho-Akt (top panels), phospho-p44/42 Erk (middle panels), or CD31 (bottom panels) antibodies to reveal activation of Akt, p44/42 Erk, and CD31⁺ vessel infiltration, respectively. Immunohistochemical staining was performed as described in Materials and Methods. The images shown are representative of sections obtained from at least five tumor samples. Results are shown for siSrc C1.